Research On The Algorithms Of Moving Object Tracking In Video

Moving object tracking in video is the critical technology of the machine vision aswell as the hot issue within certain research domains, which is widely applied insurveillance, robotics, medical field, military vision guidance, etc. However, how toaccomplish the task of robust tracking under the condition of the fast motion, occlusion,object deformation, illumination variation, background clutters, real-time restriction, etc isconcerned by scholars. It is also a tough problem yet to be resolved in practical applicationat present.In terms of the actual needs, this paper has deeply looked into the algorithms of meanshift and particle filter among many object tracking algorithms. The algorithm of meanshift is a kind of non-parametric methods based upon the climbing gradient. It realizes theaim of object tracking through iteration. The obvious mean shift is superior in less amountof calculation, simpler and easier to utilize, so it can meet the need of real-time tracking.But it fails in tracking fast moving targets and recovering a track after a severe occlusion.The third chapter gives the deduction and the description of the algorithm of mean shifttheory and its use in object tracking in detail. Meanwhile, the tracking performance of themean shift algorithm, based on color histogram, declines rapidly when the object is similarto the background color or the illumination changes. This paper focuses on the mean shifttracking algorithm on the basis of histograms of oriented gradients. The experimentalresults show that the above method is not sensitive enough to illumination changes or thepartial small deformation；Finally, the experimental analysis of drawbacks of the algorithmof mean shift is stated in the third chapter.The method of particle filters has been developed since the1990s as a new filtercalculation and applied successfully in a variety of nonlinear and non-Gaussian filteringestimations. The fundamental principle of this technique is to describe the probabilitydensities by sets of random samples, which allows online, real-time estimation of nonlinear,non-Gaussian dynamic systems. However, two common problems of the particle filter-technique are the degeneracy phenomenon and the huge computational cost. Thus,those problems will be the bottlenecks to the application of particle filter in real-timetracking systems. The algorithm of particle filter theory and its utilization in objecttracking are fully discussed in the forth chapter. The experimental results show that thealgorithm of particle filter is much more remarkable in anti-shelter and anti-inference, incomparison with the algorithm of mean shift, in addition to its worse real-time due to thehuge amount of calculation.Since object tracking algorithm based on histograms of oriented gradients alwaysloses objects under the condition of occlusion or fast motion and particle filter trackingalgorithm costs plenty of time in calculation, the object tracking algorithm based onhistograms of oriented gradients integrated with the particle filter is proposed in the fifthchapter. Under normal circumstances, the object is tracked by the algorithm based onhistograms of oriented gradient, when the conformability of candidate object is less thanthe threshold. The tracking result would be verified by particle filter algorithm. Theexperiments demonstrates that the algorithm effectively resolves the problem of theobject-loss under the condition of occlusion or fast motion and the algorithm comes outwith a better real time quality, which also overcomes the particle-degeneracy problem andachieves good tracking results under low-contrast conditions.